Maximum Entropy Multi-Task Inverse RL
This work addresses the challenge of learning from experts who may switch between multiple tasks or strategies, with applications in robotics and automation, though it appears incremental as it builds on existing maximum entropy and clustering techniques.
The paper tackles the problem of multi-task inverse reinforcement learning (IRL) by proposing a new method that combines maximum entropy IRL with Dirichlet process clustering, enabling learning of multiple reward functions from expert demonstrations. The method was evaluated on a robotic onion sorting task, achieving high accuracy and improving on previous approaches.
Multi-task IRL allows for the possibility that the expert could be switching between multiple ways of solving the same problem, or interleaving demonstrations of multiple tasks. The learner aims to learn the multiple reward functions that guide these ways of solving the problem. We present a new method for multi-task IRL that generalizes the well-known maximum entropy approach to IRL by combining it with the Dirichlet process based clustering of the observed input. This yields a single nonlinear optimization problem, called MaxEnt Multi-task IRL, which can be solved using the Lagrangian relaxation and gradient descent methods. We evaluate MaxEnt Multi-task IRL in simulation on the robotic task of sorting onions on a processing line where the expert utilizes multiple ways of detecting and removing blemished onions. The method is able to learn the underlying reward functions to a high level of accuracy and it improves on the previous approaches to multi-task IRL.